Abstract: A surface-emitting distributed feedback (SEDFB) laser is provided with a “fan-shaped” grating comprising a group of generally straight lines radiating from a central point that is away from the center of the gain. Such a grating provides better stability against self-induced filamentation and dynamic instabilities that limit achievable beam quality, especially in lasers having a stripe length greater than 1 mm.

Abstract: An electrochromic device is achieved that exhibits the characteristics of impact-resistant safety glass by subjecting a solid electrolyte sheet material and a peripheral sealant material sandwiched between substrates to heat and pressure such that the electrolyte bonds to the surfaces of the substrates with an adhesion of at least 1.8 kg/linear cm width causing the electrolyte to exhibit a tensile strength of at least 5 kg/cm2.

Abstract: A multimode source, such as a high-power laser diode bar, to pump an Nd3+ doped region defined in a cavity of a monolithic crystal structure. The axial length L1 of the doped region is chosen to optimize energy absorption from the multimode source while minimizing resonant re-absorption loss to unpumped Nd3+ ions. The next proximal cavity length L2 is an undoped region whose length is chosen to optimize the lowest order or fundamental spatial mode (“mode 9”) of the cavity. Advantageously, multi-parameter numerical optimization techniques may be employed in which the parameter set (e.g., doped length, L1, doping concentration, pump beam spot size, micro laser cavity length, and output coupler reflectivity) is varied to determine the overall optimal length L1opt.

Abstract: A prismatic device for reforming the cross section of the beam emitted along a semiconductor laser bar stripe by deflecting successive segments of the emitted beam in directions making acute angles with the longitudinal axis of the emitted beam and redirecting the deflected beam segments through layers of glass into an array of parallel beam segments.

Abstract: The mode and wavelength of a conventional, multi-mode, wide-stripe laser diode is converted in an Nd3+ ion-doped laser host crystal to a stable, single-mode output that falls within the spectral region required for pumping EDFA amplifying structures. The host crystal absorbs radiation from the diode that corresponds with the 4I9/2→4F5/2 absorption band of its Nd3+ ions, and re-radiates into the 4F3/2→4I9/2, 4F3/2→4I11/2, and 4F3/2→4I13/2 transitions which release photon energy that can be utilized by amplifying structures doped with Er3+ and/or Yb3+. The spatial mode of the multimode laser diode is converted to single-mode by enclosing the Nd3+-doped host laser crystal within a laser cavity that has a fundamental mode size large enough to encircle the spatial extent of the beam emanating from the laser diode.

Abstract: A compact wavelength division demultiplexer device comprises two cascaded interferometers the wedge angle of whose mirrors are orthogonal to each other, i.e., the two interferometers are oriented such that the dispersion of light with wavelength from the second interferometer is orthogonal to the dispersion of light with wavelength from the first interferometer. The incoming beam of laser light carrying n sub-bands of m information channels is collimated and applied to the first interferometer whose free spectral range is determined by the total bandwidth required by the n×m channels and whose finesse is determined by the resolution needed to spatially separate each of the n sub-bands from one another. The light leaving the first interferometer enters the second interferometer whose wedge angle is at 90 degrees to the plane of the wedge angle of the first interferometer.

Abstract: An image acquisition system employs a spectral imaging system whose focusing optics provide a linear dispersion of color along a tilted focal plane which intercepts each spectral component at the correct focal distance for that color.

Abstract: A high speed optical receiver comprises a photonic detector having an active aperture area dimensioned to minimize capacitance effects commensurate with achieving the desired signaling speed, and a compound parabolic reflector having a surface contour described by rotating a parabolic arc about a rotational axis, the axis of said parabolic arc making an angle with the rotational axis proportional to the conical angle of the incident light to be gathered, the reflector having a plurality of focii defining the perimeter of an exit pupil not exceeding the area of said photonic detector aperture. The compound reflector contour may advantageously be formed at the end of a length of optical fiber that is positioned at the detector aperture, incident light being admitted to the other end of the fiber through a focusing element which may be a lens, mirror, or holographic phase mask.

Abstract: A stack of parallel but laterally separated laser beams emitted from a pair of laterally separated laser diode bars are focused into a single vertical plane through the use of an interleaved array of stacked, angular glass plates. Successive plates oppositely refract laser beams from successive radiating levels of the stacks of laser bars so that beams emerging from the glass plates lie in the same vertical plane, thereby overcoming the lateral displacement among beams emerging from the laser bars.

Abstract: A group of diode lasers emitting beams at a first wavelength is positioned above and below a stack of laser diodes emitting beams at a second wavelength. The first wavelength beams are directed to a dichroic portion of a compound prism and are reflected outwardly. The second wavelength beams are directed to the opposite surface of the dichroic portion but pass through without reflection and so are directed into substantially the same outward plane as the first wavelength beams. The compound prism combines beams of two different wavelengths to achieve a greater power density in the given plane than can be obtained with beams of either wavelength alone.

Abstract: A wafer supporting a semiconductor structure having a material gain function that would preferentially support an F-P laser mode at an unwanted wavelength &lgr;2 is provided with a second order dielectric grating located sufficiently remotely from the high intensity optical field of the quantum well and the waveguide to receive just enough transverse mode energy to provide feedback to reduce the gain at &lgr;2 and support oscillation at a desired wavelength &lgr;1. More particularly, by providing a gain discrimination factor &agr;≈0.1 cm−1, the fraction of power lost to transverse mode radiation can be held to about 1% which is sufficient to provide stabilizing feedback without sapping too much energy from the longitudinal beam.

Abstract: A compact and portable docking station for a radio mobile personal digital assistant (PDA) carries a magnetic card reader and provides an interface that supplies drive power to the magnetic card reader independently of the PDA battery and translates signal levels provided from the card reader so that they can reliably be read by the PDA. PDA battery power is conserved by initiating all interface actions from a software generated “radio” button appearing on the screen of the PDA.

Abstract: An optical circulator is connected to each end of an integrated circuit chip containing a pair of multiplexer/demuliplexers driving respective arrays of 2×2 MZIs separated by a striped mirror interface and a plate having a thickness of ¼ the central wavelength on each side of the mirror. Light passes through the device twice making it polarization insensitive in power and in wavelength. Phase errors due to inserting these plates will not cause power disturbances because no interference exists. Because the same router is used for multiplexing/demultiplexing, there are no loss/crosstalk penalties for through-channels due to mismatch in wavelength response. Back reflections for the switch-conditions are suppressed by aligning the waveguide, which passes the striped mirror, under a slight angle so that reflected light at the interface will penetrate the substrate. Lateral offset applied between the ends of two waveguides at the mirror interface ensures that crosstalk performance is not limited.

Abstract: When two star couplers are cascaded so as to perform two Fourier transformations without phase distortions, an imaging arrangement results which accurately reproduces at the output the input distribution. In order to achieve high efficiency of power transfer between a relatively large number of input ports and a relatively large number of output ports and a small star-coupler physical size, the input and output waveguides connected to the star coupler must be relatively narrow and be closely spaced at the star coupler. However close spacing gives rise to significant mutual coupling between adjacent waveguides, leading to undesirable crosstalk between the channels of the device. We have discovered that the phase distortion is approximately periodic and may be compensated for by adding or subtracting length to the waveguides between the star couplers.

Abstract: A transparent chromogenic panel in which color changes are selectively effectable over predefined areas comprises a pair of facing glass substrates each covered with a conductive layer divided into individual energizeable areas each provided with as set of busbars. An electrochromic electrode layer overlies at least one of the conductive layers. An insulating adhesive sealant spaces apart the substrates and insulates the busbar sets from each other and from exposure to the electrolyte and the electrochromic layer, so that each busbar set may be individually energizeable to effect a color change through a respective one of the individual areas. A passive layer is advantageously superimposed over one of the substrates, its color being chosen so that the color and the transmissivity of the passive layer accommodates the range of color change and transmissivity of the electrochromic layer to maintain the transmitted color of the panel in a warm or neutral shade.

Abstract: Quantization noise, introduced into data transmission when analog signals are translated into PCM code using the logarithmic conversion of &mgr;-law or A-law rules necessary to accommodate the transmission of voice signals, limits the maximum attainable speed of data transmission. However, when the PCM code signals represent data (rather than voice signals), linear conversion of analog data signals into the PCM code would avoid such logarithmic quantization noise. To signal the translating codec that a digital modem call is being made, the digital modem allocates one or more of the least significant bit positions of the code representing the 2100 Hz answer-tone-with-phase-reversal (specified in ITU-T Recommendation G. 165) to send a repetitive pattern “P1” to signal to the associated codec that a digital modem connection has been made.

Abstract: A four-quadrant mixer is disclosed which has a low noise factor. The indeterminate common-mode voltage that may accompany the modulating signal is suppressed and replaced by a common-mode quiescent voltage designed to establish a predetermined quiescent biasing current through the mixer transistors common biasing resistors so that the mixer may be driven in common-mode by the modulating signal and differentially by the local oscillator signal. Advantageously, a larger value emitter biasing resistor can be used with the same value of emitter current that would obtain in a comparable four-quadrant Gilbert Mixer or, conversely, larger values of emitter current can be specified to establish a desired level of signal to noise ratio.

Abstract: A known training sequence of binary, pseudo random signals, PN, is received over the channel and correlated with a modified pseudo random sequence PN−1 in which the zeros of the known sequence PN have been mapped to −1. A transform of the correlator output will be observed to contain strong tones that indicate the degree of frequency offset. To facilitate synchronization in the presence of an unknown amount of frequency offset, the received signals may be rectified and the correlation step performed with the absolute value of the incoming signal to obtain new output samples. When the frequency offset is small its measurement is facilitated by modulating the incoming signals with a known, comparatively large frequency offset, before the correlation step to make sure that the overall frequency offset occurs around a frequency that is an integral multiple of the symbol rate divided by the number of symbols in the pseudo random sequence.

Abstract: DRAM memory unit is tested for a series of cell faults such as: the stuck-at fault (SAF), the stuck-open fault (SOF), the transition fault (TF), the multiple address fault (MAF) as well as storage capacitor leakage, subthreshold leakage or junction leakage. Predetermined data pattern is written throughout the DRAM memory, locations in one region of the memory are “frozen” while a disturbance is created in a second region during an interval sufficient for a defective cell in the first region to lose its charge, following which the locations of the first region are sequentially read to verify if any cell has lost its data.

Abstract: A codec which normally provides impedance synthesis, filtering, suppression of receive signal echoes, and gain equalization for a plurality of line circuits under the control of a microprocessor includes an arrangement for ascertaining the actual impedance of each line that it serves and making a corresponding adjustment of the receive and transmit equalizers and other codec parameters. A series of short tones having precisely known amplitude frequencies throughout the voice band are applied to the line when the line is on-hook and during the off-hook interval before tone is returned to the subscriber. Impedance is measured by observing the echo and obtaining an estimate of the phase by performing a least squares fit during a time interval not exceeding the usual interdigital call signaling interval.